Enum DeviceFlowError 

pub enum DeviceFlowError {
    AccessDenied,
    Expired,
    IncorrectDeviceCode,
    AppMisconfigured {
        code: String,
    },
    Unexpected {
        code: String,
    },
    UntrustedVerificationUri {
        origin: String,
    },
    Transport(Error),
    Status {
        status: u16,
        stage: &'static str,
    },
    Decode {
        stage: &'static str,
        source: Error,
    },
    Malformed {
        what: &'static str,
        value: String,
    },
    Config(ConfigError),
}

Every way a login can fail.

GitHub’s error_description is deliberately not carried into any of these. It is free text from a remote party, and this crate’s redaction gate is far easier to keep true by construction than by auditing what a remote string happened to contain. The machine-readable error code is documented and exhaustive, and it is what a caller branches on anyway.

Variants

AccessDenied

The user declined. Terminal, and not a failure to retry: retrying re-prompts someone who already said no.

Expired

The device code timed out before approval. A whole new login is needed — the same code cannot be re-presented.

IncorrectDeviceCode

GitHub does not recognise the device code. A new login is needed.

AppMisconfigured

The App registration itself is wrong — device flow not enabled, or a bad client_id. No amount of retrying helps; a maintainer must fix the published App (06-migration-rollout.md, Phase 0).

Fields

code: String

Unexpected

Fields

code: String

UntrustedVerificationUri

A verification_uri that is not on the configured GitHub web host. The user’s code must only ever be typed on GitHub’s own domain.

Fields

origin: String

Transport(Error)

Status

Fields

status: u16

stage: &'static str

Decode

Fields

stage: &'static str

source: Error

Malformed

Fields

what: &'static str

value: String

Config(ConfigError)

Implementations

impl DeviceFlowError

pub fn is_retryable(&self) -> bool

Whether presenting the same login again could succeed.

The two recoverable protocol states, authorization_pending and slow_down, are PollOutcomes and never become errors at all, so this is not about them. It is about the two failures that say nothing about the login: a dropped connection and a 5xx from a gateway. The device code is still live in both cases, and the next poll can still succeed — which is why DeviceFlow::complete absorbs up to MAX_TRANSPORT_RETRIES of them rather than destroying a login over a blip.

Everything else is false, and deliberately so. A caller that retried DeviceFlowError::AccessDenied would re-prompt a user who has already refused; one that retried DeviceFlowError::Expired would present a code GitHub has already discarded. Those are the cases this method exists to keep false, and no 5xx handling may be allowed to blur them.

pub fn requires_new_login(&self) -> bool

Whether the remedy is a fresh auth login rather than a maintainer fix.

Trait Implementations

impl Debug for DeviceFlowError

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Display for DeviceFlowError

fn fmt(&self, __formatter: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Error for DeviceFlowError

fn source(&self) -> Option<&(dyn Error + 'static)>

Returns the lower-level source of this error, if any.

fn description(&self) -> &str

👎Deprecated since 1.42.0:

use the Display impl or to_string()

fn cause(&self) -> Option<&dyn Error>

👎Deprecated since 1.33.0:

replaced by Error::source, which can support downcasting

fn provide<'a>(&'a self, request: &mut Request<'a>)

🔬This is a nightly-only experimental API. (error_generic_member_access)

Provides type-based access to context intended for error reports.

impl From<ConfigError> for DeviceFlowError

fn from(source: ConfigError) -> Self

Converts to this type from the input type.